In the field of video image processing, it is often necessary to decode the video image in real-time. Conventional decoding method is to decode the video image pixel by pixel. When decoding a particular pixel in the video image, if other pixels located to the left, above-left, above, or above-right of the particular pixel are already decoded, then these pixels may be used for the prediction of the particular pixel, thereby lowering the decoding workload and increasing the decoding speed. When decoding a region of pixels, the pixels are often processed in a left-to-right order through the same line. Once one line of pixels are finished decoding, the decoding process may start from the left-most pixel in a subsequent line of pixels below the decoded line of pixels.
The conventional decoding method often has an issue related to the storing of information for the decoded pixels. If a two-dimensional storage structure is used to store all the decoded pixels, then a large cache memory may be required. If a storage structure is used to store only those pixels that are located either in a horizontal or vertical line, then there may be a problem in which some of the decoded pixels in the storage structure may be overridden and rewritten during decoding.
all arranged in accordance to at least one or more embodiments of the present disclosure.